Please click here to see my previous entry Homework on a lesson -- Assignment No 1
In Assignment No 2, I discuss a second aspect put forward by Dr M. R. Srinivasan, Member (formerly Chairman) Atomic Energy Commission in his article published in The Hindu, on 11th September 2007 under the title "A lesson in nuclear reactors", in favour of importing LWRs (in the context of the civilian nuclear co-operation agreement with USA, presently being debated vigorously in our country).
Phew!! For this assignment, it took me a long time indeed to gather all relevant information in the public domain almost entirely from the Internet, understand the science behind it and present the same in this blog.
Dr. Srinivasan writes:
"Natural uranium reactors, both Magnox (GCR) and PHWR, require to be fed with fresh fuel regularly (on a daily basis). Some spent fuel has to be taken out when the reactor is operating. The PHWRs, due to their inherent lower reactivity, have a limitation on how quickly they can restart and be loaded, after an interruption due to any fault. The LWRs do not suffer from this disability. [Emphasis mine] They can run for 15-18 months without a fuel change; the latter requires the reactor to be off line for a month or so.
I would be wary of uncritically accepting this soft sell for LWRs.
I believe that the above comparison between PHWR and LWR has been made by Dr. Srinivasan based on his earlier experience with PHWRs in India which were of small capacity (220 MWe) operating in relatively small and unstable grids. In an effort to reduce the total capital cost, the volume and quantity of heavy water in the coolant circuit was kept as low as possible and hence did not have adequate "cushion" to tide over severe load-demand fluctuations from the grid side. This is not the case with modern larger PHWR (540 MWe) as well as CANDU design which have a "pressurizer" incorporated in the primary coolant circuit which greatly enhances the plant's ability to accommodate load changes (though not to as great a degree as in a thermal power plant). In other words, modern PHWRs and CANDU designs have comparable ability as LWRs in terms of load-follow capability.
Keep in mind that fissile material in the core (limited to U-235 for this discussion) provides "positive reactivity" that is necessary for fission chain reactions to take place. On the other hand all other materials in the core, including the moderator, coolant and other structural materials are neutron absorbers to a greater or lesser degree and hence contribute to "negative reactivity". Light Water used in LWR absorbs neutrons to a much greater degree than Heavy Water in PHWR. Hence LWR is considered to be very neutron-wasteful since a significant portion of the neutrons that come out of fission is absorbed in a wasteful manner, instead of being made available to cause the next fission in the chain reaction.
For chain reactions to proceed and for power generation in the core, the positive and negative reactivities need to be carefully balanced.
Fission products, which are inevitably formed in the core (fuel) -- somewhat like formation of ash when burning coal -- are also neutron absorbers. Xe-135 is one such fission product which introduces transient negative reactivity in the core. Because of its propensity to absorb neutrons, Xe is said to be a "poison" and the negative reactivity induced by Xe is called "Xe poisoning".
Dr. Srinivasan's comment above, I think, pertains to the ability of PHWR or LWR to overcome Xe poisoning effects that manifest when changes in power production levels take place in the reactor core.
As a nuclear reactor operates, Xe is produced in the core (actually in the fuel), due to a variety of nuclear reactions including the fission process itself and also radioactive decay from its precursor nuclei which are also fission products such as Tellurium and Iodine. Xe production continues for some time even after the chain reactions have stopped (that is, when the reactor is shut down). On the other hand, Xe also gets "removed" through neutron absorption, whenever chain reactions take place (that is, when the reactor is operating). The second way in which Xe gets progressively "removed" is through its natural decay. This happens when the reactor remains in the shutdown state for a length of time (of the order of 40 or 50 hours).
When rate of production is equal to rate of its removal, the Xe concentration in the core reaches an equilibrium level. So when a reactor begins to operate from a "cold clean" condition (when there was no Xe in the core), Xe formation starts and accumulates for some time until the rate of removal through neutron (produced in the fission process) absorption equals the rate of production. It attains an "equilibrium concentration".
In PHWR as well as LWR adequate excess fuel is provided to account for the equilibrium "Xe load" at all operating power levels.
Rates of Xe production and removal are affected whenever there is a change in the power production in the core. This leads to a "Xe transient" whereby the concentration of Xe present in the core may increase or decrease with time.
For purposes of this discussion, we can broadly consider two types of transients. A relatively slower Xe transient occurs when the reactor power in changed due to changes in the turbine output demand, or when operational maneuvers are carried out such as a "reactor step back", "steam dump" and "hot stand by" conditions. In all these case, the chain reactions are not stopped (that is the reactor is not tripped/scrammed). These manoeuvres lead to Xe transients whereby the Xe concentration could temporarily exceed the equilibrium concentration, before settling down to a new equilibrium value. As mentioned earlier, PHWR design incorporates features comparable to that of a LWR in successfully negotiating this transient without needing a plant shutdown.
From the point of view of this discussion on Dr. Srinivasan's comment about a so-called "disability" in the PHWR / CANDU, the more relevant type of Xe transient is one that takes place when the reactor is tripped/scrammed by introducing into the core, a massive amount of negative reactivity in a very short time, usually within about 2 seconds. Such a trip or scram may be automatically initiated by the plant control system or manually initiated by the operator whenever any one of several vital parameters (such as, for example, the coolant pressure) go beyond their (predetermined) safe values.
Figure 1 (below) is a graph depicting the Xe transient. It shows the equilibrium Xe concentration (marked "A") prior to reactor trip. The subsequent curves (in magenta and violet) show how the Xe concentration builds up and reaches a peak (marked "C") following a reactor trip. This happens because one of the two mechanisms of removal of Xe, that is, the neutron absorption process is not present subsequent to a reactor trip. After some time, production of Xe from radioactive decay of I-135 comes down while its own decay continues. So the concentration begins to go down and ultimately a level "D" is reached.
In the figure, it may be noted that the peak Xe concentration ("C") can be several times that of the equilibrium value ("A"). Usually in PHWR as well as in LWR an optimum level of extra positive reactivity is proved in the core so that some Xe poisoning ("B") can be overcome.
If for any reason the fault that caused a reactor trip/scram cannot be rectified within the "override" time, then, reactor restart would have to wait until the Xe concentration comes down to such a level that the positive reactivity that can be made available is adequate to overcome the negative reactivity due to Xe poisoning. Note that removal of negative reactivity is the same as addition of positive reactivity; so withdrawal of control/adjuster rods, or removal of deliberately added soluble poison such a Boron in the moderator has the same effect as adding positive reactivity.
In modern PHWR as well as in LWR reactivity provided to "override" Xe poisoning subsequent to a trip/scram is such that it must be restarted with in about 30 to 45 minutes. In fact, if a land-based power LWR for electricity generation were to be designed to have capability to add positive reactivity as and when required to overcome Xe poisoning to the level of "C", (this is called "full Xe override capability") then it would be even more neutron-wasteful than it already is. Only nuclear submarines which use very highly enriched fuel, have full Xe override capability as otherwise they cannot be "caught napping" while submerged at sea with their nuclear power plant in a tripped/scrammed condition even for a short while, leave alone 60 hours.
- Nuclear reactor, steam turbine, electricity generator and the grid all have to function in a co-ordinated manner for safe and successful generation, distribution and utilisation of electricity.
- Indian electricity grids, especially during the early days of nuclear power generation, tended to be small and the users often ill-behaved leading to wide fluctuations in voltage and frequency. Such steep variations in operating parameters, usually led the nuclear power plant to trip and separate from the grid as a protective measure to avoid damages to the turbo-machinery.
Click image to enlarge (in new window) - As shown in Figure 2, modern nuclear power plants (both PHWRs and LWRs) have the ability to succesfully decouple the turbo generator from the grid and also the reactor from the turbogenerator so that in most cases a grid fault or certain malfunctions in the turbogenerator systems do not require a reactor trip/scram. In these conditions the reactor can be “islanded” and continued operation at a lower power output is made possible by supplying “house loads” and/or by steam discharge; the reactor stays critical (that is, it is not tripped/scrammed).
- Reactor design takes into account the above situation and as long as the reactor is not tripped/scrammed, enough excess reactivity can be made available to overcome the build up of Xe “poisoning” that might temporaily result from the reduction in the output.
- This is true for both LWRs as well as PHWRs.
- On the other hand, should a fault occur in any of the systems associated with the nuclear reactor plant itself (or in the turbogenerator system) that requires a reactor trip/scram, then, it must be cleared within about 30 to 45 minutes to be able to restart the reactor before Xe poison build up takes place to such an extent that it cannot be overcome. If this window of opportunity is lost (that is if the fault causing the reactor trip could not be rectified within 30 to 45 minutes), then plant operators must wait until the Xe poison dies down sufficiently (about 60 hours) before the reactor can be made critical once again.
- Again (6) above is true for LWRs as well as PHWRs although plant-specific information on the exact time duration available to the operator to “beat the Xe poison” could not be readily obtained in the public domain from the Internet. Only submarine reactors which use highly enriched U235 have full Xe override capability and can be restarted at any time after a reactor trip. In a land-based electricity generating nuclear power plant, to provide for full Xe override capability by building a large amount of excess reactivity in the core would lead to a significant reduction in neutron economy.
- Modern large capacity PHWR / CANDU is more neutron-economical than a LWR and has similar operational capabilities. To me it seems that imported LWRs do not enjoy a significant advantage over PHWRs (such as India-built Tarapur 3&4) with regard to Xe override capability as indicated by Dr. Srinivasan.
- What is more important is the overall availability/capacity factor of the nuclear power plant. According to NPCIL's Annual Report for 2006-07 weighted availability factor for its power plants was 85%. Figure 3 below demonstrates that performance of Indian PHWRs (indicated in the figure as NPCIL PHWRs) are comparable to foreign plants.
Click image to enlarge (in new window) - It is not just the quality of the design that governs successful operation of a power plant. Good operations and management culture combined with sufficient understanding of know-why is more important than just having imported systems/equipment. Over the last few decades India has acquired much of the all important know-why, the hard way. It is essential this is not lost by resorting to imported equipment/systems. On the contrary, it must be enhanced by providing all the appropriate financial, governmental, managerial and consistent policy support.
